Since their discovery (by Secchi 1867) until the end of the sixties the Be stars, challenging and all-time variable emission-line stars, have not been the subject of any systematic study aimed at detection of their possible light and colour variations. It is true that the astronomical literature of that period contains ample evidence that many Be stars are light variables but almost all such findings resulted as by-products of studies of different or wider groups of objects.
The first discovery of the light variations of Be stars goes back to the times of Johannes Kepler (P Cyg). Feinstein (1968) was probably the first who pointed out explicitly that many Be stars are light variables. Perhaps the first study explicitly aimed at the detection of light variations of a large group of Be stars by means of differential photoelectric photometry was published by Haupt & Schroll (1974). The interest in systematic photometric studies of Be stars has been growing since then.
The history of discoveries of the light variability of Be stars and the new developments of this field in the seventies and early eighties has already been summarized in several reviews (cf., e.g., Harmanec 1983a,b; Percy 1987; Cuypers 1991) and need not be repeated here. More recent surveys of the photometric variability of Be stars were published, for instance, by Schuster & Alvarez (1983), Schuster & Guichard (1984), Percy et al. (1988b), Cuypers et al. (1989), Balona et al. (1992) or Percy & Attard (1992). A systematic program of uvby long-term monitoring of Be stars was initiated by Sterken (1983) within the framework of the ESO long-term monitoring program and the results are being published in a series of papers (Mennickent et al. 1994; Sterken et al. 1996).
In 1972, the Hvar Observatory was opened. Systematic photoelectric UBV observations of bright Be stars has become the main observing program of the photometric telescope of the Observatory. The original goal of the program was to search for new eclipsing binaries among them. However, already the first results attracted the interest of observers also to long-term and later to rapid variations of Be stars.
In realization that a systematic investigation of Be variability cannot be carried out from one observatory only, an international campaign on systematic photoelectric observations of all bright Be stars was organized and coordinated under the auspices of the Working Group on Be Stars of the IAU Com. 29 by Harmanec et al. (1980b, 1981a,b, 1982a,b). Since the aim of the program
was to accumulate statistically significant data on the Be phenomenon without any prejudice, the list of objects (with obligatory comparisons) included all emission-line B stars of all luminosity classes and - for most part - brighter than the magnitude limit of the Bright Star Catalogue.
The Be program at Hvar has become an important part of this effort. We tried to observe all groups of variables defined in the program which were observable from Hvar. In practice, however, not all of the selected objects were intensively or are at all observed that location, for one reason or another. Observations of only a few emission-line stars of luminosity classes Ia-II were secured, for instance.
Many of the findings which resulted from the Be program have been subsequently published in a series of papers devoted to individual Be stars (see Table below), in Progress Reports in the Be Star Newsletters Nos. 2 to 10 and partly also in the following summary reports: Harmanec et al. (1980a), Koubský & Pavlovski (1982), Pavlovski & Bož ić (1982), and Harmanec (1983a,b, 1987b, 1994).
The principal finding of our studies was that the long-term light variations of Be stars are usually the most pronounced ones and that they are somehow related to the long-term spectral variations of respective stars. While we failed to discover new eclipsing binaries among Be stars, we did find phase-locked periodic light variations for several known Be binaries (KX And, LQ And, RX Cas, V1507 Cyg, CX Dra, V360 Lac, Per, Tau). The data from Hvar, supplemented with nearly simultaneous observations from other stations were also used, especially for studies of rapid light variations. The "minicampaign" organized in 1983 (see Stagg et al. 1988) may serve as a good example.
The publication of this study was substantially delayed after we had realized that the simple bilinear formulae which we used did not ensure sufficiently-stable seasonal transformations to the standard UBV system and can even lead to false reports of light variability if, for instance, an A or F star is used as a comparison for an early B star, or for reddened objects. Only the recent development of new reduction procedures (based on nonlinear transformation equations) and a careful determination of more accurate UBV magnitudes of all comparison and standard stars used in the Be program (Harmanec et al. 1994a) opened the way to completion of this study. In a preliminary form, an overview of the UBV photometry of Be stars at Hvar was given by Pavlovski et al.\ (1994).